• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Maximum Likelihood Estimation of Mathematical Models for Genetic Development in Gastrointestinal Stromal Tumors

Posted on: Wednesday, 15 September 2004, 06:00 CDT

367

Aims: Cytogenetic and molecular data of gastrointestinal stromal tumors (GIST) indicate consistent patterns of genetic progression, including KIT activating mutation and subsequent accumulation of chromosomal aberrations. Reconstruction of the temporal sequence of chromosomal changes may help to further our understanding of the critical events involved in the genetic progression of GIST. We herein present the first oncogenetic models for karyotypic development in GIST.

Methods: The study is based on a genetic data set of 170 GIST obtained by comparative genomic hybridization (CGH). Maximum likelihood estimation was applied to analyze the sequential order of chromosomal imbalances with the aim of inferring typical pathways of karyotypic evolution. On this basis, probabilistic mathematical models were generated that model the likely sequences of occurrence of non-random chromosomal imbalances.

Results: A general order of sequences for several DNA copy number changes emerged, allowing the identification of early, intermediate or late events in genetic tumor progression. Distinct pathways of karyotypic evolution and the relationship of different pathways in GIST could be identified.

Conclusions: The identified tree models of likely sequences of acquired DNA copy number changes reflect the multistep process, that is required beyond initiating KIT mutations for the development and progression of GIST.

A. VON HEYDEBRECK, B. GUNAWAN1, D. KOVAC2, M. BOLLMANN3 M. VINGRON, L. FZESI1

Max-Planck-Institut fr Molekulare Genetik, Berlin

1 Institut fr Pathologie, Universitt Gttingen

2 Institut fr Pathologie, Universitt Rijeka, Kroatien

3 Institut fr Pathologie, Bonn

Copyright Urban & Fischer Verlag 2004

More News in this Category